This invention relates to a frequency-division multiplexing transceiver apparatus and method for sending and receiving data by a mobile-station-specific frequency spectrum, and more particularly to a frequency-division multiplexing transceiver apparatus and method for transmitting a transmission symbol upon subjecting the symbol to phase rotation that varies at a speed specific to the mobile station.
DS-CDMA (Direct Sequence-Code Division Multiple Access) multiplies a narrow-band transmission signal by a spreading code in order to spread and transmit that transmission signal over a wider band. In DS-CDMA, when each of a plurality of mobile stations sends a transmission signal upon multiplying it by a spreading code having a certain spreading factor SF, the information transmission speed becomes 1/SF. Therefore, in order to achieve a frequency utilization efficiency that is equivalent to that of TDMA, it is necessary in DS-CDMA to accommodate a number of signals that is equal to SF number of mobile stations. However, in an actual wireless propagation environment on the uplink, the effect of Multiple Access Interference (MAI), in which the signals from each of the mobile stations interfere with each other, becomes dominant due to differences in propagation conditions from each mobile station to the base station, for example, due to differences in propagation-delay time or propagation-path fluctuation, and thus the rate of frequency utilization decreases.
Therefore, IFDMA (Interleaved Frequency Division Multiple Access) is being studied as a wireless modulation method that is capable of reducing the effects of MAI in next-generation mobile communications (see the specification of JP2004-297756 A, and “Investigations on Packet Error Rate of Variable Spreading and Chip Repetition Factors (VSCRF)-CDMA Wireless Access in Reverse Link Multi-cell Environment”, The Institute of Electronics, Information and Communication Engineers Technical Report of IEICE, RCS2004-84 (204-206)). This IFDMA modulation method transmits a transmission signal upon multiplying the signal by a phase that changes at a speed specific to the mobile station, thereby reducing MAI by placing the signals from each of the mobile stations on a frequency axis in such a manner that the signals will not overlap each other on the frequency axis.
FIG. 21 is a block diagram showing the structure of a mobile station that uses an IFDMA modulation method, and FIG. 22 is a drawing that explains an IFDMA symbol. A channel encoder 1a performs channel encoding by applying error-correction encoding such as turbo encoding or convolutional encoding to an entered binary information sequence, and a data modulator 1b converts the channel-encoded data to I, Q complex components (symbols) in QPSK, for example. A symbol transmitted in one frame of IFDMA is referred to as an “IFDMA symbol”, and one IFDMA symbol is composed of Q-number of symbols S0, S1, S2, S3 as shown in (a) of FIG. 22 (Q=4 in the figure).
A symbol-repetition and rearrangement unit 1c compresses the time domains of the four symbols S0, S1, S2 and S3 of the IFDMA symbol, and repeatedly generates each symbol L times (L=4 in the figure), as well as rearranges the repeatedly generated symbols and places them in the same arrangement as that of the symbol sequence S0, S1, S2, S3 (see (b) of FIG. 22). By taking Tc to be the sample period, the period Ts of symbol-repetition will satisfy the relation Ts=Tc×Q. A phase-rotation unit 1d has a complex multiplier CML that performs mobile-station specific phase rotation of each symbol in the repetitive symbol sequence (see (c) of FIG. 22), and a wireless transmitter 1e performs up-conversion of the signal that is input from the phase-rotation unit 1d from baseband frequency to radio frequency, after which it amplifies the signal and transmits it from an antenna.
When the time domains of the transmission-symbol sequence S0, S1, S2, S3 are compressed and each transmission symbol is repeatedly generated a prescribed number of times (L times), and each of the symbols of the repetitive-symbol sequence are rearranged so as to have the same arrangement as that of the symbol sequence S0, S1, S2, S3, the repetitive-symbol sequence after rearrangement will have a comb-tooth-shaped frequency spectrum as shown in (a) of FIG. 23. Also, by performing phase rotation that varies at a speed that is specific to the mobile station of each of the symbols of the rearranged repetitive-symbol sequence, the spectral positions of the comb-tooth-shaped frequency spectrum shift as shown in (a) to (d) of FIG. 23, and frequency-division multiplex transmission becomes possible. In other words, when the speed of phase rotation is zero, the frequency spectrum of the output signal from the phase-rotation unit 1d will have comb-tooth-shaped frequency spectrum characteristics as shown in (a) of FIG. 23, and as the amount of change in the phase rotation per unit time Tc (frequency) increases, the frequency spectrum will shift as shown in (a) to (d) of FIG. 23.
An NCO (Numerically Controlled Oscillator) 1g calculates the amount of phase rotation θ per unit time Tc, and the complex multiplier of the phase-rotation unit 1d performs phase rotation specific to the mobile station for each symbol of the repetitive-symbol sequence and executes frequency shift processing.
The phase θk(t) that is output from the NCO 1g after repeating Q-number of symbols L times is given by the following equation:
                                                        θ              k                        ⁡                          (              t              )                                =                                                    k                ·                2                            ⁢              π              ⁢                                                          ⁢                                                W                  L                                ·                t                                      =                                          k                ·                2                            ⁢              π              ⁢                                                          ⁢                                                1                                      L                    ·                    Q                    ·                    Tc                                                  ·                t                                                    ⁢                                  ⁢                  W          =                      1            Ts                          ⁢                                  ⁢                  QW          =                      1            Tc                                              (        1        )            where W is the symbol frequency, and k is a value that corresponds to the mobile station and is any one value among 0, 1, 2, . . . L−1. NCO 1g outputs the phase θk(t) as θ, which has been calculated according to Equation (1), at the period Tc, and is so adapted that the amount of phase rotation will be 2π at the IFDMA period (=L·Q·Tc=16 Tc) (such that the phase will make one full cycle).
In NCO 1g, a frequency-shift-setting unit 1h sets the amount of change of phase rotation per unit time Tc (angular speed) Δω, and using the parameters k, L and Q, calculates the angular speed Δω according to the following equations:
                                          Δ            ⁢                                                  ⁢            ω                    =                                                    k                ·                2                            ⁢              π              ⁢                                                          ⁢                              W                L                                      =                                          k                ·                2                            ⁢              π              ⁢                                                          ⁢                              1                                  L                  ·                  Q                                                                    ⁢                                  ⁢                  f          =                                                    Δ                ⁢                                                                  ⁢                ω                                            2                ⁢                                  π                  ·                  Tc                                                      =                          k                              L                ·                Q                ·                Tc                                                                        (        2        )            A rotation-phase-amount-setting unit 1i comprises an adder ADD and a delay unit DLY for applying a delay time T (=Tc), and performs a calculation according to the following equation every unit time Tc to increase the rotation phase θ by Δω at a time and output the result.θ=θ+Δω  (3)A converter 1j calculates I, Q components (x, y) in a complex plane of the rotation phase-amount θ and inputs these components to the phase-rotation unit 1d. By taking the symbols of the repetitive-symbol sequence to be S (=X+jY), the phase-rotation unit 1d performs a calculation according to the following equation and outputs the calculation result.(X+jY)·(x+jy)In actuality, the complex multiplier CML of the phase-rotation unit 1d calculates and outputs (Xx−Yy) and (Xy+Yx) for each real-number and imaginary-number part.
If k=0, the amount of change in phase rotation will be zero (Δω=0), and the frequency spectrum will become as shown in (a) of FIG. 23. If k=1, the amount of change in phase rotation will become Δω=2π/L×Q according to Equation (2), and if Q=L=4, then the phase will change in increments of π/8 as shown in (c) of FIG. 24, and the frequency spectrum will become as shown in (d) of FIG. 24 or (b) of FIG. 23. Also, if k=2, the amount of change in phase rotation will become Δω=4π/L×Q according to Equation (2). If Q=L=4, then the phase will change in increments of π/4 for each Tc, and the frequency spectrum will become as shown in (c) of FIG. 23. Moreover, if k=3, then the amount of change in phase rotation will become Δω=6π/L×Q according to Equation (2). If Q=L=4, then the phase will change in increments of 3π/8 for each Tc, and the frequency spectrum will become as shown in (d) of FIG. 23. As a result, even when a plurality of mobile stations access the same base station simultaneously, the frequency spectrum of each mobile station will be orthogonal on the frequency axis, and it is possible to reduce interference among transmission signals.
FIG. 25 is a block diagram showing different structure of a mobile station that uses an IFDMA modulation method, and FIG. 26 is a drawing that explains an IFDMA symbol. The mobile station shown in FIG. 25 spreads the transmission symbols using a spreading code, and the phase of the chip-repetitive sequence that is obtained by compressing and repeating the time domains of the spread chip sequence obtained by spreading is rotated at each unit time Tc.
A channel encoder 1a performs channel encoding by applying error-correction encoding such as turbo encoding or convolutional encoding to an entered binary information sequence, and a data modulator 1b converts the channel-encoded data to I, Q complex components (symbols) in QPSK, for example. One IFDMA symbol is composed of Q-number of symbols as shown in (a) of FIG. 26 (Q=2 in the figure).
A spreading-code-multiplier 1m multiplies the symbols S0, S1 with the spreading code c00, c01, c10, c11 of the spreading factor (SF=2 in the figure) to generate a spread chip sequence (see (b) of FIG. 26). As a result, one IFDMA symbol is composed of Q×SF (=4) chips C00, C01, C10, C11.
A chip-repetition unit 1n compresses the time domains of the four chips C00, C01, C10, C11 of the spread chip sequence, and repeats each chip CRF times (CRF=4 in the figure), then rearranges the chip-repetitive sequence so that it has the same arrangement as the original chip sequence C00, C01, C10, C11 (see (c) and (d) of FIG. 26). Here, CRF is an abbreviation for Chip Repetition Factor, Taking Tc to be the chip repetition period, the period Ts of the repetitive-chip sequence becomes Ts=Tc×Q×SF.
A phase-rotation unit 1d has a complex multiplier CML that performs mobile-station specific phase rotation of the chip-repetitive sequence (see (e) of FIG. 26), and a wireless transmitter 1e performs up-conversion of the signal that is input from the phase-rotation unit 1d from baseband frequency to radio frequency, after which it amplifies the signal and transmits it from an antenna.
When the speed of phase rotation is zero, the frequency spectrum of the signal output from the phase-rotation unit 1d has the frequency-spectrum characteristics as shown in (a) of FIG. 23, and as the amount of change in phase rotation per unit time Tc (frequency) becomes large, the frequency spectrum shifts as shown in (a) to (d) of FIG. 23.
An NCO (Numerically Controlled Oscillator) 1g calculates the amount of phase rotation θ per unit time Tc, and the complex multiplier CML of the phase-rotation unit 1d performs phase rotation specific to the mobile station for each chip of the repetitive-chip sequence and executes frequency shift processing.
The phase θk(t) that is output from the NCO 1g is given by the following equation:
                                                        θ              k                        ⁡                          (              t              )                                =                                                    k                ·                2                            ⁢              π              ⁢                                                          ⁢                                                W                  CRF                                ·                t                                      =                                          k                ·                2                            ⁢              π              ⁢                                                          ⁢                                                1                                      CRF                    ·                    SF                    ·                    Q                    ·                    Tc                                                  ·                t                                                    ⁢                                  ⁢                  W          =                      1            Ts                          ⁢                                  ⁢                              SF            ·            Q            ·            W                    =                      1            Tc                                              (        4        )            where k is a value that corresponds to the mobile station and is any one value among 0, 1, 2, . . . CRF−1. NCO 1g outputs the phase θk(t) which has been calculated according to Equation (4), at the period Tc, and is so adapted that the amount of phase rotation will be 2π at the IFDMA period (=16 Tc) (such that the phase will make one full cycle).
In NCO 1g, a frequency-shift-setting unit 1h sets the amount of change of phase rotation per unit time Tc (angular speed) Δω, and using the parameters k, CRF, Q and SF calculates the angular speed Δω according to the following equations:
                                          Δ            ⁢                                                  ⁢            ω                    =                                                    k                ·                2                            ⁢              π              ⁢                                                          ⁢                              W                CRF                                      =                                          k                ·                2                            ⁢              π              ⁢                                                          ⁢                              1                                  CRF                  ·                  SF                  ·                  Q                                                                    ⁢                                  ⁢                  f          =                                    Δω                              2                ⁢                π                ⁢                                                                  ⁢                Tc                                      =                          k                              CRF                ·                SF                ·                Q                ·                Tc                                                                        (        5        )            A rotation-phase-amount-setting unit 1i comprises an adder ADD and a delay unit DLY for applying a delay time T (=Tc), and performs a calculation according to the following equation every unit time Tc to increase the rotation phase θ by Δω at a time and outputs the result.θ=θ+Δω  (6)A converter 1j calculates I, Q components (x, y) in a complex plane of the rotation phase amount θ and inputs these components to the phase-rotation unit 1d. The phase-rotation unit 1d performs phase rotation specific to the mobile station for each chip of the repetitive-chip sequence, and performs frequency-shift processing.
If k=0, the amount of change in phase rotation will be zero (Δω=0), and the frequency spectrum will become as shown in (a) of FIG. 23. If k=1, the amount of change in phase rotation will become Δω=2π/CRF×SF×Q according to Equation (5), and if Q=SF=2, and CRF=4, then the phase will change in increments of π/8 and the frequency spectrum will become as shown (b) of FIG. 23. Also, if k=2, the amount of change in phase rotation will become Δω=4π/CRF×SF×Q according to Equation (5). If Q=SF=2 and CRF=4, then the phase will change in increments of π/4 for each Tc, and the frequency spectrum will become as shown in (c) of FIG. 23. Moreover, if k=3, then the amount of change in phase rotation will become Δω=6π/CRF×SF×Q according to Equation (5). If Q=SF=2 and CRF=4, then the phase will change in increments of 3π/8 for each Tc, and the frequency spectrum will become as shown in (d) of FIG. 23. As a result, even when a plurality of mobile stations access the same base station simultaneously, the frequency spectrum of each mobile station will be orthogonal on the frequency axis, and it is possible to reduce interference among transmission signals.
A mobile station is normally battery operated, and in order to lengthen the possible communication time, it is desired that the efficiency of the transmission amplifier in the wireless unit be increased. Also, in order that the transmission signal is not distorted by the transmission amplifier, linear characteristics are desired. The input/output characteristics of the transmission amplifier show linear characteristics when the input power is low as shown in FIG. 27, however, as the input power increases, the input/output characteristics become non-linear when the input power becomes Pmax or greater, as shown by the dotted line. When the input power becomes Pmax or greater, non-linear distortion occurs due to this non-linear characteristic.
As operation point approaches Pmax, in order to use the transmission amplifier with high efficiency, the input power exceeds Pmax and distortion occurs when the transmission signal is large, however, when operation point is decreased so that distortion does not occur, the efficiency of the transmission amplifier drops. In order to satisfy these conflicting needs, it is important that the PAPR (Peak to Average Power Ratio) of the transmission signal be made small. PAPR is the ratio between the peak power value and average power value of the transmission signal. When PAPR is large, the signal that is output from the amplifier becomes distorted when the operation point of the amplifier approaches Pmax and the transmission signal is at a peak, and when PAPR is small, the signal that is output from the amplifier does not become distorted when the transmission signal is at a peak, and it is possible to use the amplifier efficiently.
There is prior art whose object is to reduce the peak factor in the CDMA modulation method (see the specification of JP2005-57582 A). In this prior art, when a multiplexed CDMA signal exceeds the input-limit of the transmission amplifier, power control is performed for data symbol near a specified symbol location in I-Q complex plane, and power-level control is performed, taking into consideration the symbol rate of the user and required SIR.
However, in this prior art, the PAPR of the transmission signal is not reduced in the IFDMA modulation method.
Technical Report of IEICE, RCS2004-84 (204-206)
[Non-patent Document 1] Goto, et al., “Investigations on Packet Error Rate of Variable Spreading and Chip Repetition Factors (VSCRF)-CDMA Wireless Access in Reverse Link Multi-cell Environment”, The Institute of Electronics, Information and Communication Engineers,